Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

use fdtbus_intr_establish_xname

branches: 1.1.2; 1.1.6; 1.1.16;
Add generic AHCI SATA controller bus glue

Add amdsmn(4) and amdccp(4) power management stubs.

branches: 1.6.18;
Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

Use DEVICE_COMPAT_EOL.

Since we're using designated initialisers for compat data, we should
use a completely empty initializer for the sentinel.

Remove "struct of_compat_data" and replace its usage with
"struct device_compatible_entry"; they are ABI-compatible.

Fix several "loses const qualifier" bugs encountered during
this conversion.

trailing whitespace

branches: 1.1.2; 1.1.6; 1.1.16;
Add amdccp(4) driver for AMD Cryptographic Coprocessor, as found on the
A11xx Opterons. Driver currently provides RNG service only.

fix RK3399 emmc and set SDHC_FLAG_NO_PWR0.

tested on rockpro64 and pinebookpro.

Re-enable ADMA2 support for arasan,sdhci-8.9a

Disable ADMA2 on the Arasan SDHCI 8.9A found in the Xilinx Zinq-7000 due
to sporadic transfer errors until the root cause for the errors is found.
In the meantime, SDMA works fine on this platform.

No need to constrain DMA tag to 32-bits on 32-bit platforms.

Use generic Arasan SDHCI driver for Zynq-7000.

sdhc: Retire SDHC_FLAG_USE_ADMA2 flag.

ADMA2 support in sdhc is mature now, so no need for it to be opt-in.

Use __MASK(4) in bus_dmatag_subregion. NFCI.

s/0xffffffff/UINT32_MAX/ in bus_dmatag_subregion call

Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

Remove "struct of_compat_data" and replace its usage with
"struct device_compatible_entry"; they are ABI-compatible.

Fix several "loses const qualifier" bugs encountered during
this conversion.

use fdtbus_intr_establish_xname

branches: 1.3.10;
Add SDHC_FLAG_SINGLE_POWER_WRITE and SDHC_FLAG_32BIT_ACCESS flags

Enforce 32-bit limits on ADMA2 capable controllers that do not support 64-bit descriptors

branches: 1.1.4;
Add support for Arasan SDHCI 5.1

Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

branches: 1.5.8;
Set sysclk rate at set_format time, so the link set_format callback can read the new sysclk

branches: 1.4.2;
Merge isaki-audio2 branch, the overhaul of audio subsystem.
- Interrupt-oriented system rather than thread-oriented.
- Improve stability, quality and performance.
- Split playback and record cleanly. Improve halfduplex support.
- Many bugs are fixed including deadlocks, resource leaks, abuses, etc.
- Simplify audio filter mechanism. The encoding/channels/frequency
conversions are completely handled in the upper layer. So the hard-
ware driver only converts its hardware encoding (if necessary).
- audio_hw_if changes:
- Obsoletes query_encoding and add query_format instead.
- Obsoletes set_params and add set_format instead.
- Remove drain, setfd, mappage.
- The call sequences are changed.
- ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted.
- ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced.
- cleanup config attributes: au*conv and mulaw.
- All hardware drivers should follow it (I've done as much as possible).

Some file paths are changed:
- dev/audio.c -> dev/audio/audio.c (rewritten)
- dev/audiovar.h -> dev/audio/audiovar.h
- dev/audio_dai.h -> dev/audio/audio_dai.h
- dev/audio_if.h -> dev/audio/audio_if.h
- dev/audiobell.c -> dev/audio/audiobell.c
- dev/audiobellvar.h -> dev/audio/audiobellvar.h
- dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c

branches: 1.3.2; 1.3.4; 1.3.8;
Pass mixer requests to the codec. It is the responsibility of the codec
to decide whether or not to dispatch to an aux device.

Attach aux devices to the codec

Add SoC sound driver based on "simple-audio-card" DT binding spec.

risc-v: attach the Cadence XHCI usb controller on the JH7100 SoC

As with ACPI, perform the fdtbus_register_i2c_controller() in a centralized
location.

Garbage-collect fdtbus_attach_i2cbus(); the regular iicbus_attach() is
sufficient now.

Add driver for Cadence I2C controller.

Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

Use DEVICE_COMPAT_EOL.

Since we're using designated initialisers for compat data, we should
use a completely empty initializer for the sentinel.

Remove "struct of_compat_data" and replace its usage with
"struct device_compatible_entry"; they are ABI-compatible.

Fix several "loses const qualifier" bugs encountered during
this conversion.

branches: 1.1.2; 1.1.18;
Add connector and panel drivers (panel supports only panel-lvds and
panel-dual-lvds at this time, but can easily be extended to other types
of panels).
Add an API for ports/endpoints.
Proposed on tech-kern@ a few days ago, ok jmcneill@

branches: 1.1.2;
Add connector and panel drivers (panel supports only panel-lvds and
panel-dual-lvds at this time, but can easily be extended to other types
of panels).
Add an API for ports/endpoints.
Proposed on tech-kern@ a few days ago, ok jmcneill@

Re-factor the operating point defintions into fdt_opp.h

It is more appropriate to use clk_get_rate() rather than sc->sc_freq_target * 1000000.
ci_data.cpu_cc_freq should be set to a higher precision value.

In addition, when cpufreq_dt_init(), or while throttling, sc->sc_freq_target should not
be referenced by cpufreq_dt_change_cb() because it does not have the correct value.

Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

branches: 1.17.2;
Re-enable "set cpufreq at boot" code.

Disable setting cpufreq at boot for now

Highest speed is first opp, not last

Select the highest rate at boot, and update cycle counter frequency when rates change

Add explicit FDT_OPP for operating-points-v2 so the link set won't be empty

Add support for platform specific opp table filters.

Skip nodes with an "opp-suspend" property and fix tables that have disabled
nodes in the middle.

Fix argument to fdtbus_get_reg, from Jared.

Change sysctl to be named after the first CPU in the DVFS domain.

old: machdep.cpu.frequency.*, machdep.cpufreqdt4.frequency.*
new: machdep.cpufreq.cpu0.*, machdep.cpufreq.cpu4.*

branches: 1.8.2;
opp-v2 table seems to list opps from lowest to highest, so swap the order to ensure the highest frequency is first

Support for "opp-microvolt" with more than one cell

Use machdep.cpu for the first instance

Don't print an error message for shared opp tables

Add support for opp-v2 tables and multiple instances of the cpufreqdt
driver.

branches: 1.3.2; 1.3.4;
Support cpu nodes without a cpu-supply property.

branches: 1.2.2;
Listen for PMFE_THROTTLE_* events and limit CPU frequency when throttling
is enabled.

Add generic DT cpu frequency and voltage scaling driver.

Call / define fdtbus_cpus_md_attach for platforms with cpus @ fdt.

The RISC-V binding here seems somewhat of an abuse, but it exists in
mainline linux.

RISC-V support that works on QEMU with a single hart.

Thanks for Simon Burge for plic(4).

Fix inverted result when status is "okay"

Do not sort cpu nodes when enumerating so they attach in the order listed
in the devicetree.

Attach cpu nodes with status "disabled" if they have an enable-method
property. This is a valid configuration according to the devicetree
specification.

cpus: use fdt_add_bus

branches: 1.2.6; 1.2.10; 1.2.12; 1.2.14;
Allow devices to attach to cpu devices

Add CPUs ("/cpus" node) driver. CPUs is a container of cpu nodes.

branches: 1.1.6; 1.1.10;
Add helper for parsing display timings.

Multiple inclusion protection define consistency

branches: 1.1.6; 1.1.10; 1.1.28;
Add helper for parsing display timings.

Re-factor the console-related code into fdt_console.[ch]

fix various typos, mainly in comments, but also log messages, docs, game text.

dw_apb_uart: Honour reg-io-width property

Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

use fdtbus_intr_establish_xname

branches: 1.8.2;
Revert rev 1.5:
http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/dev/fdt/dw_apb_uart.c#rev1.5

The device is capable to recognize break signal actually.
Reset cnmagic from +++++ to default.

Pointed out by jakllsch. Thanks!

Default reg-shift is 2, not 0, for DW APB UART

Use com_init_regs_stride instead of a4x tag

branches: 1.5.2;
The device cannot recognize break signal. Use +++++ (five plus signs) as
cnmagic in the same manner with bcm2835_com.c.

Clean up initialization of com_regs structure, in preparation for
some additional changers.

Use PRI?64 instead of ll?

branches: 1.2.2; 1.2.4;
Enable apb_pclk clock if present

Rename sunxi_com to dw_apb_uart and move it to dev/fdt since it is not
sunxi-specific.

dw2c_fdt: Terminate statements by ; instead of , (from Mori Hiroki)
No binary changes.

Merge thorpej-cfargs2.

branches: 1.11.8;
Merge thorpej-cfargs branch:

Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.

Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)

Remove unnecessary or redundant interface attributes where they're not
needed.

There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)

...and a sentinel value CFARG_EOL.

Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.

Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).

branches: 1.10.2;
Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

Use DEVICE_COMPAT_EOL.

Since we're using designated initialisers for compat data, we should
use a completely empty initializer for the sentinel.

Remove "struct of_compat_data" and replace its usage with
"struct device_compatible_entry"; they are ABI-compatible.

Fix several "loses const qualifier" bugs encountered during
this conversion.

use fdtbus_intr_establish_xname

branches: 1.5.12;
Add support for Meson GXBB

Add support for Meson8b

Remove useless header arm/broadcom/bcm2835reg.h

From Aymeric Vincent on port-arm

branches: 1.2.2; 1.2.4;
Match generic snps,dwc2 compat string

Add initial support for Rockchip RK3328 SoC.

Attach when dr_mode is "otg" and the controller has the "usb-role-switch" and
"role-switch-default-mode" as seen in the Apple M1 dts.

This assumes the controller is properly setup for host mode already.

Look for child node by compat string snps,dwc3 instead of by name.

Remove "no IOMMU" hack.

Until we have proper FDT iommu support, refuse to attach when an iommu
is required.

Merge thorpej-cfargs2.

branches: 1.15.4;
Unbreak previous; we have to look for a child node iff our node isn't
compatible with snps,dwc3.

branches: 1.14.2; 1.14.4;
Merge thorpej-cfargs branch:

Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.

Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)

Remove unnecessary or redundant interface attributes where they're not
needed.

There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)

...and a sentinel value CFARG_EOL.

Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.

Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).

branches: 1.13.2;
Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

use fdtbus_intr_establish_xname

branches: 1.11.2;
Initialise xhci_softc sc_ios

Add "snps,dwc3" to the list of compatible strings we match against.
The code already handles "snps,dwc3" not being a subordinate of
an SoC-specific node, but the string was missing from the match
routine.

Necessitated by the sun50i-h6 device tree update on Jan 3 2020, which
elimiated the "allwinner,sun50i-h6-dwc3" node and placed all of the
clocks, resets, etc. directly under the "snps,dwc3" node.

Add support for NXP i.MX 8M Dual/8M QuadLite/8M Quad family SoCs.

branches: 1.8.2;
Add support for snps,dis-tx-ipgap-linecheck-quirk quirk

branches: 1.7.4;
Add support for Meson GXL USB3 controller

Add support for RK3399

Add support for Samsung Exynos USB3 DRD.

Fix name of snps,dis_u2_susphy_quirk property

Detect USB 3.1.

branches: 1.2.2;
Add support for RK3328

branches: 1.1.2;
Add DesignWare DWC3 XHCI driver.

Match "starfive,jh7110-mmc"

branches: 1.23.2;
Restrict dwcmmc to 32bit DMA (<4GB) regardless of attachment

Limit DMA to below 4GB - the supported controllers are limited this
way.

Thanks to jmcneill for a version of this patch.

spaces to tabs

dwcmmc_fdt_bus_clock: fix an aprint_debug_dev frequency units botch.

fdt: dwcmmc: Honour broken-cd and non-removable DT properties

Fix previous

Match generic "snps,dw-mshc" compat string as found in BeagleV.

branches: 1.16.2;
Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

Use DEVICE_COMPAT_EOL.

Since we're using designated initialisers for compat data, we should
use a completely empty initializer for the sentinel.

Remove "struct of_compat_data" and replace its usage with
"struct device_compatible_entry"; they are ABI-compatible.

Fix several "loses const qualifier" bugs encountered during
this conversion.

use fdtbus_intr_establish_xname

branches: 1.11.6;
dwc_mmc fixes:
- Rockchip uses a different SDIO int bit, so take this into consideration
- Avoid unnecessary resets and always wait for resets to complete
- kpause instead of delay while holding spinlock
- Do not attempt autostop for SD_IO_RW_EXTENDED commands
- Allow for sub-blklen byte counts for single block transfers

branches: 1.10.2;
Fix performance regression with previous

dwcmmc improvements:
- Use mmcpwrseq resources if available
- Only set 4- or 8-bit mode if specified in the dt properties
- Add quirk for implementations with inverted power enable logic
- Support switching signal voltage between 1.8V and 3.3V
- Fix a clock divider issue on Rockchip SoCs

branches: 1.8.2;
Use correct size of softc in CFATTACH_DECL_NEW

Use more generic rk3288 compatible string, and do clock assignments when present

Use PRI?64 instead of ll?

sc_clock_freq is in Hz, not kHz

Fix 32-bit build

branches: 1.3.2; 1.3.4;
If max-frequency is not specified, use UINT_MAX instead of the rate programmed by the bootloader.

dwcmmc: set ciu clock rate rather than assuming fixed input rate

Add initial support for Rockchip RK3328 SoC.

Separate DesignWare watchdog driver and FDT glue.

Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

branches: 1.4.8;
dwcwdt: make this work correctly

- sysmon_wdog.smw_period is seconds, not milliseconds
- tickle the watchdog before enabling it

branches: 1.3.4;
Deactivate the watchdog timer during attach.

Recent u-boot starts up the watchdog timer, and we don't detect this so
e.g. my DE0 nanosoc reboots after about 20 seconds. Applying a reset to
the watchdog circuit is the only way to stop it according to the
documentation.

Make compile

branches: 1.1.2;
Add driver for DesignWare Watchdog Timer. Untested as I can't seem to find
the parent clock for this in the RK3328 documentation.

As with ACPI, perform the fdtbus_register_i2c_controller() in a centralized
location.

Garbage-collect fdtbus_attach_i2cbus(); the regular iicbus_attach() is
sufficient now.

Fix previous with respect to the binding specification which I misread.
There must be 1 clocks (but there can be two) and the reset is optional.

dwiic_fdt: enable the clock and de-assert the reset

Improve some error handling while I'm here.

branches: 1.5.8;
dwiic(4): Don't try to attach children if dwiic_attach failed.

PR kern/57063

branches: 1.4.4; 1.4.14;
Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

use fdtbus_intr_establish_xname

Change fdtbus_register_i2c_controller() to directly register the i2c_tag_t,
rather than the device and a set of functions (the only of which was to
return the i2c_tag_t anyway). Previously, this assumed only a single
i2c controller node per device_t, which is not true with an i2c mux.

branches: 1.1.2; 1.1.6; 1.1.16;
Add dwiic_fdt attachment for "snps,designware-i2c".

Merge thorpej-cfargs2.

branches: 1.7.8;
Merge thorpej-cfargs branch:

Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.

Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)

Remove unnecessary or redundant interface attributes where they're not
needed.

There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)

...and a sentinel value CFARG_EOL.

Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.

Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).

branches: 1.6.2;
Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

use fdtbus_intr_establish_xname

branches: 1.4.14;
Register shutdown handlers

branches: 1.3.2;
Stop potential misuse of vendor names and USB vendor IDs in root hub
device and string descriptors.

Firstly: Few vendors have identical PCI-SIG vendor IDs and USB-IF vendor
IDs. As such, using the PCI vendor ID as a USB vendor ID may trample
on whomever is allocated that USB vendor ID.

Secondly: The vendor of the host controller hardware implementation has
little to nothing to do with our usbroothub implementation. Thus we
should not potentially associate any problems therewith to such third
party.

This change will result in root hubs being identified by USB Vendor ID
0x0000. Root hub vendor string will now be "NetBSD" (or, specifically:
ostype). Product ID (0x0000) and product strings remain unchanged.

branches: 1.2.4; 1.2.6; 1.2.8; 1.2.10;
Remove the hack to find companion devices and just assume 1 companion if
ETTF flag is not set.

Add glue for generic ehci/ohci bindings.

Allocate enough address space for unaligned start/end values.

branches: 1.6.2;

Include <sys/bootblock.h> for struct mbr_sector

Make fdt_cpu_rootconf available for other machines/platforms.

G/C some #if 0 / #endif code

fdt(4): Factor out bootargs support from evbarm and riscv.

Factor out some fdt(4) features from {,evb}arm into dev/fdt.

Now, FDT-based support to efirt, initrd, rndseed, and efirng can be
used from, e.g., riscv.

Mostly from Nick Hudson.

XXX
As Nick comments, there can be some optimizations for fdt_map_range().
efiboot may also be modified to load these objects into aligned PAs.

Move fdt_update_stdout_path from evbarm code to MI code

Make fdt_cpu_rootconf available for other machines/platforms.

fdt(4): Factor out bootargs support from evbarm and riscv.

Factor out some fdt(4) features from {,evb}arm into dev/fdt.

Now, FDT-based support to efirt, initrd, rndseed, and efirng can be
used from, e.g., riscv.

Mostly from Nick Hudson.

XXX
As Nick comments, there can be some optimizations for fdt_map_range().
efiboot may also be modified to load these objects into aligned PAs.

Move fdt_update_stdout_path from evbarm code to MI code

Add fdtbus_clock_enable and fdtbus_clock_enable_index shortcuts

Add fdtbus_clock_count to count the number of clock references on a given node

branches: 1.8.4;
Consolidate FDT [find "names" index by string] logic.

Adjust residual calculation in "find index by name" idioms to take into
account the trailing NUL on each name. Some other similar instances of
this idiom already account for this.

(In preparation to factor out this idiom into its own function.)

Pass clock provider's phandle to fdtbus_clock_controller_func.decode()
and update callers.

This allows to accomodate clock managers whose clocks are identified
directly by a clock instead of a pair (clock provider, index).

ok jmcneill@ on port-arm

Use queue(3) API to manage lists. NFCI.

branches: 1.4.2;
fdtbus_clock_assign: simplify

Process assigned clock parents and rates on clock provider nodes.

Add fdtbus_clock_byname, which can be used by clock backends to
lookup clocks in other domains by "clock-output-names" property.
Not intended for ordinary driver use.

branches: 1.1.2; 1.1.18; 1.1.20;
Add support for fdt clock and reset controllers.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Re-factor the console-related code into fdt_console.[ch]

Re-factor the console-related code into fdt_console.[ch]

Use queue(3) API to manage lists. NFCI.

branches: 1.2.2;
Remove unnecessary include

branches: 1.1.2;
Add SoC sound driver based on "simple-audio-card" DT binding spec.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Improve fdt_isprint so that it returns false if there are consecutive
zeroes.

branches: 1.1.2;
Add a "show fdt" ddb command

Multiple inclusion protection define consistency

branches: 1.1.2;
Add a "show fdt" ddb command

Consolidate FDT [find "names" index by string] logic.

Adjust residual calculation in "find index by name" idioms to take into
account the trailing NUL on each name. Some other similar instances of
this idiom already account for this.

(In preparation to factor out this idiom into its own function.)

Use queue(3) API to manage lists. NFCI.

branches: 1.1.2; 1.1.8; 1.1.12; 1.1.14; 1.1.16;
Add FDT DMA controller API.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Add fdtbus_gpio_count(), which counts the number of GPIO entries
in a specified property.

branches: 1.6.14;
Use queue(3) API to manage lists. NFCI.

branches: 1.5.2; 1.5.4; 1.5.6;
Add fdtbus_gpio_acquire_index for accessing multi-xref gpios properties.

Memory leak, found by mootja; not tested, but obvious enough

branches: 1.3.2; 1.3.4;
Add fdtbus_gpio_{read,write}_raw, which tells the controller not to take
polarity into account. Tegra GPIO pin data includes pin polarity, but so
does a regulator-fixed node, so the end result was that the enable value
was being swapped twice. Change fregulator to use the raw APIs, and adapt
Tegra and Exynos GPIO drivers to support this flag.

Only OF_* functions should be in fdt_openfirm.c, move the rest to fdt_subr.c.

split gpio, i2c, intr, and regulator helpers into separate files

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

G/C fdtbus_copy_i2c_devs() and directly enumerate the devices using
the device tree in the FDT case.

As with ACPI, perform the fdtbus_register_i2c_controller() in a centralized
location.

Garbage-collect fdtbus_attach_i2cbus(); the regular iicbus_attach() is
sufficient now.

As with ACPI, iic_attach() now fetches the child device array directly
in the FDT case. Adjust the OFW case so that we're not forced to indirect
through a dictionary if it's not needed.

API naming consistency. NFC.

branches: 1.11.2;
Merge thorpej-cfargs2.

branches: 1.10.2; 1.10.8;
Merge thorpej-cfargs branch:

Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.

Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)

Remove unnecessary or redundant interface attributes where they're not
needed.

There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)

...and a sentinel value CFARG_EOL.

Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.

Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).

branches: 1.9.2;
Change fdtbus_register_i2c_controller() to directly register the i2c_tag_t,
rather than the device and a set of functions (the only of which was to
return the i2c_tag_t anyway). Previously, this assumed only a single
i2c controller node per device_t, which is not true with an i2c mux.

branches: 1.8.2;
i2c-indirect-config == false --> i2c-no-indirect-config == true

Fix leak of iba.iba_child_devices in fdtbus_attach_i2cbus(). Patch
provided by yarl-baudig %%at%% mailoo.org

Add support for Allwinner A64's display pipeline.

Set "i2c-indirect-config" to false in the right place so that iic(4)
better behaves itself.

Add fdtbus_attach_i2cbus, used to attach an i2cbus to an fdt device node.

Use queue(3) API to manage lists. NFCI.

branches: 1.2.2; 1.2.18; 1.2.20; 1.2.22;
Only OF_* functions should be in fdt_openfirm.c, move the rest to fdt_subr.c.

split gpio, i2c, intr, and regulator helpers into separate files

G/C fdtbus_copy_i2c_devs() and directly enumerate the devices using
the device tree in the FDT case.

As with ACPI, perform the fdtbus_register_i2c_controller() in a centralized
location.

Garbage-collect fdtbus_attach_i2cbus(); the regular iicbus_attach() is
sufficient now.

As with ACPI, iic_attach() now fetches the child device array directly
in the FDT case. Adjust the OFW case so that we're not forced to indirect
through a dictionary if it's not needed.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

fdt: add helper for finding intr parent phandle

Add 'const char *xname' param to fdtbus_intr_establish_byname

add fdtbus_intr_establish_xname() function

Add 'const char *xname' arg to fdtbus_interrupt_controller_func .establish

branches: 1.26.4;
Use MUTEX_DEFAULT to avoid confusion

Add fdt_intr_mask() / fdt_intr_unmask() functions, for masking and
unmasking invididual interrupt sources (similar to acpi_intr_mask()
and acpi_intr_unmask()).

branches: 1.24.2;
Remove a stupid printf

be more verbose about errors.

Add support "interrupts-extended".

* fdtbus_get_phandle_with_data().
Add utility subroutine to get phandle with data.

Add fdtbus_intr_establish_byname() helper.

Avoid using the "interrupt-map" table on our own interrupts.

This is needed by the rk3399.dtsi's pcie0 node, which needs to map its
interrupts an entity other than itself.

Suggested by jmcneill@

Fix sense of test in fdtbus_intr_disestablish.

Reported by yarl-baudig on port-arm.

Add fdtbus_intr_establish_raw and fdtbus_intr_str_raw, for establishing
interrupts directly using an interrupt controller's phandle and specifier.

Get rid of unnecessary calls to kmem_alloc

Return the interrupt controller's cookie directly from fdtbus_intr_establish

When searching for interrupt parent, stop as soon as we see an
interrupt-controller property on a node other than the starting node.

Fix a regression with interrupt-map handling in previous.

interrupt-map bug fixes

Use queue(3) API to manage lists. NFCI.

branches: 1.11.4; 1.11.6; 1.11.8;
Fix an issue with interrupt controller lookup wrt cascading interrupt
controllers.

branches: 1.10.2;
Fix interrupt-map support on vexpress-a15 and try to avoid unnecessary
calls to kmem_alloc.

Rewrite interrupt-map support based on ePAPR spec.

Fix get_specifier_by_index to work with indexes greater than 0.

FDT Interrupt handling - change memory allocation

This version does dynamic allocation of the specifier array.

FDT interrupts - bug fixing

Fix a pair of memory leaks
Add some descriptive messaging in fail cases.
Fix that _by_index wasn't walking the tree upward to find the interrupt-parent.
Fix a pair of messages so they include the function in the text.

FDT interrupts -- clean up debug printfs

Ooops. forgot to delete my debug printfs before my last checkin.

FDT: Interrupts -- add support for interrupt maps

The mct on exynos uses an interrupt map so we add support now. Devices
represent their interrupts either through a combination of interrupt-parent
and interrupts properties, where the 'interrupts' property is an array of
one or more interrupt specifiers; or through a combination of an
interrupt-parent that points to an interrupt-map, where the interrupt-map
contains 2 or more entries consisting of an index, a pointer to an
interrupt-controller, and a specifier for that controller.

This code adds the ability to walk the interrupt-map and return a specifier.
Unfortunately, the addition requires changing the interface to the
interrupt-controllers' _establish and _intstr functions, so this check in
contains a rototill of the three existing fdt interrupt controllers to use
the new interface.

branches: 1.3.2;
Use of_getprop_bool/of_getprop_uint32

Only OF_* functions should be in fdt_openfirm.c, move the rest to fdt_subr.c.

split gpio, i2c, intr, and regulator helpers into separate files

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Add facility for registering iommus.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Add mailbox and power-domain support.

power-domain support is very rudimentary.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Handle multiple memory nodes by type, but still allow an untyped
(non-standard) node named /memory.

Cache the result from fdtbus_get_data() in fdt_memory_remove_reserved

NFCI.

Size boot_physmem with FDT_MEMORY_RANGES.

This effectively increases the size from 64 to 256 entries for Arm kernels.
It turns out on large systems that memory can be quite fragmented by UEFI.
Increasing the size of this just kicks the can down the road, we need a
better solution to deal with boot_physmem/fdt_memory/bootparams.dram
sizing.

Do what the comment says and reserve ranges that do and do not have the
"no-map" property.

Required for qemu/riscv

In my previous change, it was necessary to consider the case where the first block size is zero.

Don't pass a block of size 0 to fdt_memory_add_range().

There are some environments where size 0 blocks are passed from the loader.

Calculate the minimum address, don't assume the first entry is the start.

Fix 32-bit build.

fdt "/reserved-memory" node should be handled

branches: 1.1.2; 1.1.6;
Move evbarm/fdt/fdt_memory.[ch] to sys/dev/fdt and simplify the api and
some operations. This allows other architectures to use it.

Size boot_physmem with FDT_MEMORY_RANGES.

This effectively increases the size from 64 to 256 entries for Arm kernels.
It turns out on large systems that memory can be quite fragmented by UEFI.
Increasing the size of this just kicks the can down the road, we need a
better solution to deal with boot_physmem/fdt_memory/bootparams.dram
sizing.

branches: 1.1.2;
Move evbarm/fdt/fdt_memory.[ch] to sys/dev/fdt and simplify the api and
some operations. This allows other architectures to use it.

Use queue(3) API to manage lists. NFCI.

branches: 1.1.2; 1.1.4; 1.1.6;
Add support for simple MMC power sequence provider bindings.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

branches: 1.2.2; 1.2.18;
Only OF_* functions should be in fdt_openfirm.c, move the rest to fdt_subr.c.

Add a framework for enumerating devices using a Flattened Device Tree (FDT).

Only OF_* functions should be in fdt_openfirm.c, move the rest to fdt_subr.c.

Add a framework for enumerating devices using a Flattened Device Tree (FDT).

Re-factor the operating point defintions into fdt_opp.h

Sort includes.

Get drm to build on arm64 again.


Author: Jared McNeill <jmcneill@NetBSD.org>
Committer: Taylor R Campbell <riastradh@NetBSD.org>

dts: adapt to dts-5.15 bindings

Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

branches: 1.2.4; 1.2.8; 1.2.12;
Abort panel driver attach if required regulator is missing.

Add another panel@fdt driver, this time for DRM-style panels.

To do: migrate away from other panel driver.

Skip xref if it is 0

branches: 1.5.4;
Consolidate FDT [find "names" index by string] logic.

Adjust residual calculation in "find index by name" idioms to take into
account the trailing NUL on each name. Some other similar instances of
this idiom already account for this.

(In preparation to factor out this idiom into its own function.)

Add support for Allwinner A64's display pipeline.

Use queue(3) API to manage lists. NFCI.

branches: 1.1.4; 1.1.6; 1.1.8; 1.1.10; 1.1.12;
Add FDT PHY interface.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Add support for devices with separate "init" and "default" pinctrl states.

Don't assume pinctrl-0 is the default configuration.

branches: 1.8.4;
Consolidate FDT [find "names" index by string] logic.

Implement subroutines for parsing out some of the generic properties
specified in the pinctrl bindings, and adapt Meson, Rockchip, and
Allwinner pinctrl back-ends to use them.

Ok jmcneill@

Fix infinite loop like in fdtbus_get_string_index.

Use queue(3) API to manage lists. NFCI.

branches: 1.4.4; 1.4.6; 1.4.8;
Fix the pinctrl api to match the spec. A pinctrl config can have more
than one xref, and an xref may have specifier data associated with it.

branches: 1.3.8;
Memory leak, found by mootja; not tested, but obvious enough. By the way,
I guess we should be handling the return value of OF_getprop.

branches: 1.2.2; 1.2.4;
FDT pinctl - review from Jared

These changes reflect a redesign based on a preliminary review by Jared.
Instead of the acquire/release/set/get approach of the original, this uses
a much simpler, and cleaner register/set approach.

FDT pinctrl

Add a pinctrl bus to FDT. This works against exynos, but someone(tm) needs
to think about whether it is general enough or too specific to exynos.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

G/C dummy FDT_PLATFORM to fix NULL-dereference crash during boot

observed for official test suite runs for aarch64 and riscv64.

This became apparent with link-order change due to FDT refactoring.
Need to pull up to netbsd-11.

Refactor the "platform" defitions into fdt_platform.h

branches: 1.1.8;
Rename ARM_PLATFORM to FDT_PLATFORM and make it available outside arm.

Refactor the "platform" defitions into fdt_platform.h

when enumerating ports and endpoints treat missing 'reg' properties as zero
ok jmcneill:
Looking at Linux. If port or endpoint are missing a 'reg' property it defaults to 0.
Please make our code do the same.

see https://github.com/devicetree-org/dt-schema/blob/main/dtschema/schemas/graph.yaml

with this my pinebook has a usable console again

Quiet chatty printfs

style fix/KNF

Fix build

Use fdtbus_get_reg to read "reg" property

branches: 1.2.4;
Add support for Allwinner A64's display pipeline.

branches: 1.1.2; 1.1.4;
Add connector and panel drivers (panel supports only panel-lvds and
panel-dual-lvds at this time, but can easily be extended to other types
of panels).
Add an API for ports/endpoints.
Proposed on tech-kern@ a few days ago, ok jmcneill@

fix various typos in comments and log messages.

fix various typos in comment, documentation and log messages.

fix various typos in comments and output/log messages.

Multiple inclusion protection define consistency

branches: 1.3.12;
Add EP_DRM_PANEL endpoint type

Add support for Allwinner A64's display pipeline.

branches: 1.1.2; 1.1.4;
Add connector and panel drivers (panel supports only panel-lvds and
panel-dual-lvds at this time, but can easily be extended to other types
of panels).
Add an API for ports/endpoints.
Proposed on tech-kern@ a few days ago, ok jmcneill@

Use queue(3) API to manage lists. NFCI.

branches: 1.1.6; 1.1.10; 1.1.12; 1.1.14;
Add facility for fdt devices to register reset and poweroff handlers.
Use this to remove as3722-specific code in tegra_machdep.c

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

fdt_powerdomain: Fix bug by which pd index >= 1 couldn't be enabled

Length in bytes was mistakenly used as number of uint32_t variables.

branches: 1.1.4;
Add mailbox and power-domain support.

power-domain support is very rudimentary.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

branches: 1.1.2; 1.1.6;
Add fdt_intr_mask() / fdt_intr_unmask() functions, for masking and
unmasking invididual interrupt sources (similar to acpi_intr_mask()
and acpi_intr_unmask()).

(This file was missed in a previous commit.)

Use queue(3) API to manage lists. NFCI.

branches: 1.1.2; 1.1.4;
Add support for PWM backlights.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

fdt: regulator: pre-allocate regulator handle

Workaround for PR# port-evbarm/54664

Honour regulator-enable-ramp-delay

Add a function to report if a regulator can support the requested voltage range

Use queue(3) API to manage lists. NFCI.

branches: 1.5.2;
I guess fdtbus_regulator_get_voltage() should check for the get_voltage()
callback, not set_voltage()

branches: 1.4.2; 1.4.4;
Print an error if we fail to acquire a regulator

Add regulator APIs for setting and getting voltage.

branches: 1.2.2; 1.2.4;
Only OF_* functions should be in fdt_openfirm.c, move the rest to fdt_subr.c.

split gpio, i2c, intr, and regulator helpers into separate files

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Consolidate FDT [find "names" index by string] logic.

Adjust residual calculation in "find index by name" idioms to take into
account the trailing NUL on each name. Some other similar instances of
this idiom already account for this.

(In preparation to factor out this idiom into its own function.)

Use queue(3) API to manage lists. NFCI.

branches: 1.1.2; 1.1.18; 1.1.20; 1.1.22;
Add support for fdt clock and reset controllers.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Garbage-collect fdtbus_todr_attach(); todr_attach() does all the
necessary work now.

kern/59630

Introduce a device call "device-is-system-todr", which allows the
platform device tree to be consulted if there are multiple TODR / RTC
devices present on the system; if the call is not supported, the behavior
is un-changed: the first TODR device is attached, all others ignored. If
the call is supported, devices other than the specified system TODR are
skipped.

Implement this device call for FDT.

kern/59630

branches: 1.1.2; 1.1.10; 1.1.14;
Add fdtbus_todr_attach.

The kernel will only ever use the first RTC driver to call todr_attach.
When drivers use fdtbus_todr_attach, if an "rtc0" alias exists in the
devicetree, ensure that only that device node calls todr_attach.

In spi_configure() attempt to fetch transfer mode information from the
platform device tree. Implement this for FDT per the SPI controller and
periperal Device Tree bindings.

Make SPI direct configuration less #ifdef'y. No more checking what kind
of platform device tree; instead, use a device call to enumerate the
direct config children of the bus.

Garbage-collect fdtbus_attach_spibus(). spibus_attach() does everything
now.

Put the task of getting the device array from the FDT onto the
SPI bus attach routine. This is slightly ugly now, but a prettier
solution is on the way, and this at least funnels the ugliness into
a single location.

Do the #address-cells dance inside of_copy_spi_devs(), and make calling
it a bit simpler.

Remove pointless containment of SPI device array inside a dictionary
that never gets passed on to anyone.

- Const poison references to spi_controller.
- Don't make a copy of the spi_controller in the SPI bus softc; just take
a reference to the one we got from the back-end driver.

Register the SPI controller with FDT in spi_attach(), not in each
driver.

It is not necessary to pass the phandle separately to fdtbus_spi_*(); it
can be retrieved from the device_t.

Don't bother registering a function that returns a SPI controller; just
register the controller directly.

branches: 1.3.2;
Merge thorpej-cfargs2.

branches: 1.2.2; 1.2.8;
Merge thorpej-cfargs branch:

Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.

Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)

Remove unnecessary or redundant interface attributes where they're not
needed.

There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)

...and a sentinel value CFARG_EOL.

Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.

Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).

branches: 1.1.8; 1.1.12;
fdt: add SPI controller frontend

Garbage-collect fdtbus_attach_spibus(). spibus_attach() does everything
now.

- Const poison references to spi_controller.
- Don't make a copy of the spi_controller in the SPI bus softc; just take
a reference to the one we got from the back-end driver.

It is not necessary to pass the phandle separately to fdtbus_spi_*(); it
can be retrieved from the device_t.

Don't bother registering a function that returns a SPI controller; just
register the controller directly.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

Step towards modularizing the Flattened Device Tree code.

Define attributes for each of the specific device bindings: clock,
dai, dma, gpio, i2c, iommu, mbox, mmc_pwrseq, phy, power, power domain,
pwm, regulator, reset controller, spi, system controller, pin
controller. Include these support files only if either a provider
or consumer with one of these attributes is present in the kernel
config.

Add the necessary attributes to the device / attach declarations for
each provider and consumer.

There are some bindings that are consumed by generic code (iommu, pinctrl,
power, power domain). Provide weak stubs for these routines to handle
situations where there is no provider.

No actual code changed; NFCI.

Re-factor the console-related code into fdt_console.[ch]

G/C fdt_console_list -- nothing uses it.

Use strlist(9).

branches: 1.38.2;
Add kernel option FDT_DEFAULT_STDOUT_PATH to set default (fallback)
stdout-path

Allow kernels to link with no console drivers

Fix LOCKDEBUG by calling fdtbus_intr_init as part of cpu_startup_hook, i.e.
later.

0x%#x --> %#x for non-external codes.
Also, stop mixing up 0x%x and %#x in single files as far as possible.

Add dma-ranges support.

Add fdt_intr_mask() / fdt_intr_unmask() functions, for masking and
unmasking invididual interrupt sources (similar to acpi_intr_mask()
and acpi_intr_unmask()).

Rename fdtbus_set_data() to fdtbus_init(). NFC.

branches: 1.31.2;
Use correct #cells for parent address when decoding ranges

Add support "interrupts-extended".

* fdtbus_get_phandle_with_data().
Add utility subroutine to get phandle with data.

Consolidate FDT [find "names" index by string] logic.

Adjust residual calculation in "find index by name" idioms to take into
account the trailing NUL on each name. Some other similar instances of
this idiom already account for this.

(In preparation to factor out this idiom into its own function.)

Hide address decoding debug messages under FDTBUS_DEBUG

Fix merge botches. I hope there's no more left...

Merge the [pgoyette-compat] branch

Provide an FDTBASE option which allows kernels to access FDT parsing
functions without requiring full blown FDTisation.

This will be used by ODROID-C1 in move to generic start code.

Use PRI?64 instead of ll?

branches: 1.22.2;
Fix fdtbus_get_cells(, 2) when running big endian.

fdtbus_get_string_index(): decrease len when going to the next string,
otherwise we may read past the end of names.

branches: 1.20.2;
FDTise RapberryPI support. Thanks for jmcneill for a lot of help with this.

The kernel image that the RPI firmware boots is now netbsd.img in the
kernel build directory.

XXX fdtbus_get_reg needs reworking

branches: 1.19.2;
fdtbus_get_string: support special 'name' property

Add fdtbus_get_reg_byname

Fix a problem with fdtbus_get_phandle where we were passing size -1 to
kmem_alloc when the property was not found. This seemed to work with
DIAGNOSTIC kernels, but panics in vmem with a non-DIAGNOSTIC kernel.

Add fdtbus_get_string_index helper.

Fix the pinctrl api to match the spec. A pinctrl config can have more
than one xref, and an xref may have specifier data associated with it.

Add fdtbus_get_string helper

branches: 1.13.2;
The "ranges" property provides a means of translating between the
address space of a bus and the address space of the bus node's parent.
Translate addresses using these rules in fdtbus_get_reg{,64}.

Move console initialization out of platform code into the console drivers
themselves.

Handle reg values beyond 4GB for platforms with 32-bit bus_addr_t.

fdtbus_get_reg will do a range check and return ERANGE if addr+size
exceeds 4GB.

Add fdtbus_get_reg64 to allow for fetching raw values w/o range check.

Move logic for checking "status" property out of fdtbus into a public
fdtbus_status_okay function.

branches: 1.9.2;
If the stdout-path property is missing on /chosen, assume serial0:115200n8

Add some helpers to find the stdout device.

Avoid using kmem in fdtbus_get_reg so we can use it early at boot to read
the physical memory layout of the system.

Print the full DT node path in "not configured" autoconf messages.

Before: clock at fdt2 not configured
After: /clocks/clock@0 at fdt2 not configured

branches: 1.5.2; 1.5.4; 1.5.6;
Add support for fdt clock and reset controllers.

Use of_getprop_bool/of_getprop_uint32

Only OF_* functions should be in fdt_openfirm.c, move the rest to fdt_subr.c.

split gpio, i2c, intr, and regulator helpers into separate files

Add a framework for enumerating devices using a Flattened Device Tree (FDT).

branches: 1.3.2; 1.3.8;
Simplify list initialization

Add helper to lookup syscon by phandle

Add support for system controller nodes.

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

fdt: add support for USERCONF option

Nothing uses fdt-path anymore, and it's redundant with device-path,
so g/c.

Allow the fdtbus_attach_args to fdt_add_bus() to be NULL.

Add mailbox and power-domain support.

power-domain support is very rudimentary.

Change the devhandle_from_*() functions to also take a "super handle",
from which the newly created handle will inherit it's implementation.
The root implementation for a new handle type is used if an invalid
"super handle" is passed.

fdtbus: match simple-pm-bus

Instead of requiring drivers to be explicit about their device memory
mapping requirements, move this decision into the fdt layer. This
introduces a new MD function, fdtbus_bus_tag_create, which is responsible
for returning per-node bus_space handles.

Merge thorpej-cfargs2.

branches: 1.41.8;
Merge thorpej-cfargs branch:

Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.

Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)

Remove unnecessary or redundant interface attributes where they're not
needed.

There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)

...and a sentinel value CFARG_EOL.

Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.

Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).

branches: 1.40.2;
Assign device handles to devices that attach to the "fdt" interface attribute
in fdtbus_device_register().

Add fdtbus_device_register(). Just a placeholder for now.

Rename of_match_compat_data() to of_compatible_match(). Similarly,
rename of_search_compatible() to of_compatible_lookup().

Standardize on of_compatible_match() for driver matching, and adapt
all call sites.

Use DEVICE_COMPAT_EOL.

Replace 2 uses of of_match_compatible().

branches: 1.35.2;
Adjust fdt_rescan to speed up booting by significantly reducing the number
of times that we call config_match. Now we only call it 11 times per node.
Previously, each scan pass of an unconfigured device triggered 11 calls to
config_match, and adding a new simplebus causes the scan to restart, so
we are avoiding potentially hundreds of calls to config_match per node now.

Update for proplib(3) API changes.

Add dma-ranges support.

First arg of cpu_setmodel needs to be a format string - fix the llvm build.
From maxv.

When attaching the first fdtbus, use the root "comptabile" (or failing that:
"model") property to set the cpu model (in userland aka sysctl hw.model).
When attaching the first cpu, do not overwrite a cpu model if it already
had been set.

branches: 1.30.2;
Add support for devices with separate "init" and "default" pinctrl states.

branches: 1.29.2;
Restore "not configured" autoconfiguration prints by using config_found_sm_loc on the default pass

Reduce the number of times we need to call driver match functions. Makes
a noticeable difference in boot time on an ODROID-XU4.

Attach syscon to "simple-mfd" devices and fix a bug with fdt_find_with_property

Add a helper to search the tree for a node with the specified property

Do not sort cpu nodes when enumerating so they attach in the order listed
in the devicetree.

Support rescan and child detachment

Add fdt_add_bus_match, which is like fdt_add_bus but allows for the caller to filter child nodes on their own

Store node path in device properties "fdt-path" property:

# drvctl -p ehci0 fdt-path
/soc/usb@5101000

While here, remove fdt_scan_bus (no longer required) and some debug printfs.

Allow nodes to attach child devices using the same sorting rules as used
by the simplebus driver.

When attaching child devices, consider the best match for all passes
instead of just the current pass. Fixes an issue where a driver with a
lower match number and early pass number will attach instead of a better
match on a later pass.

branches: 1.19.2;
Revert previous.

In fdtbus_print(), aprint_normal the path to the device (rather than
aprint_debug). This info is every bit as useful as, say, PCI device
locations.

If we don't have a native driver for a node and it has a simple-mfd
compatible string, treat it the same as a simple-bus.

Allow to remove entries from the fdt, either by handle or
by compatible string. Proposed on tech-kern and port-arm, ok jmcneill@

branches: 1.15.2; 1.15.4;
Scan children of the /chosen node to pick up simplefb nodes

Attach fdtbus to a /clocks node with no compatible string.

branches: 1.13.2;
Change first pass from 1 to 0.

Rename fdt_print to fdtbus_print and make it public.

Move logic for checking "status" property out of fdtbus into a public
fdtbus_status_okay function.

Print node path in debug log.

branches: 1.9.2;
Allow for specifying (optional) pass number locator for devices at fdt.

Attach devices in order of "phandle" property inserted by DTC.

Change order in which nodes in the devicetree are processed.

1. Attach all known busses.
2. Attach all "early init" devices (by compatible string).
3. Finally, attach all other devices.

Remove simplebus driver and allow fdtbus to attach directly to itself.

Allow devices to attach to nodes without a "compatible" property, but only
print "not configured" messages for nodes with one.

Print the full DT node path in "not configured" autoconf messages.

Before: clock at fdt2 not configured
After: /clocks/clock@0 at fdt2 not configured

Fix non-DIAGNOSTIC build

branches: 1.2.2; 1.2.4; 1.2.6;
sometimes status is "ok" instead of "okay", allow both

Add a framework for enumerating devices using a Flattened Device Tree (FDT).

Garbage-collect fdtbus_todr_attach(); todr_attach() does all the
necessary work now.

kern/59630

Re-factor the console-related code into fdt_console.[ch]

Re-factor the operating point defintions into fdt_opp.h

G/C fdt_console_list -- nothing uses it.

Refactor the "platform" defitions into fdt_platform.h

Re-factor the FDT binding-specific definitions and function prototypes
into their own header files.

fdt: Add a fp_device_register_post_config callback to fdt_platform.

branches: 1.79.6;
Call / define fdtbus_cpus_md_attach for platforms with cpus @ fdt.

The RISC-V binding here seems somewhat of an abuse, but it exists in
mainline linux.

Rename ARM_PLATFORM to FDT_PLATFORM and make it available outside arm.

Add mailbox and power-domain support.

power-domain support is very rudimentary.

Sort a bit. NFC

API naming consistency. NFC.

Formatting. NFCI.

fdt: add helper for finding intr parent phandle

Instead of requiring drivers to be explicit about their device memory
mapping requirements, move this decision into the fdt layer. This
introduces a new MD function, fdtbus_bus_tag_create, which is responsible
for returning per-node bus_space handles.

Add facility for registering iommus.

branches: 1.70.2; 1.70.12;
Merge thorpej-cfargs branch:

Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.

Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)

Remove unnecessary or redundant interface attributes where they're not
needed.

There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)

...and a sentinel value CFARG_EOL.

Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.

Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).

branches: 1.69.2;
Add fdtbus_device_register(). Just a placeholder for now.

oops, fix mistake in previous commit. removed unwanted member that had been added.

Add 'const char *xname' param to fdtbus_intr_establish_byname

add fdtbus_intr_establish_xname() function

Add 'const char *xname' arg to fdtbus_interrupt_controller_func .establish

Change fdtbus_register_i2c_controller() to directly register the i2c_tag_t,
rather than the device and a set of functions (the only of which was to
return the i2c_tag_t anyway). Previously, this assumed only a single
i2c controller node per device_t, which is not true with an i2c mux.

Add fdtbus_gpio_count(), which counts the number of GPIO entries
in a specified property.

Multiple inclusion protection define consistency

branches: 1.61.2;
Get rid of a4x bus_space tag from fdtbus_attach_args. The only consumer
of this was various com(4) glue so modify all of that to use the new
com_init_regs_stride instead.

Add dma-ranges support.

Add fdt_intr_mask() / fdt_intr_unmask() functions, for masking and
unmasking invididual interrupt sources (similar to acpi_intr_mask()
and acpi_intr_unmask()).

Rename fdtbus_set_data() to fdtbus_init(). NFC.

branches: 1.57.2;
Add fdtbus_clock_enable and fdtbus_clock_enable_index shortcuts

Add fdtbus_clock_count to count the number of clock references on a given node

Add support for platform specific opp table filters.

Add support for devices with separate "init" and "default" pinctrl states.

fdt: add SPI controller frontend

branches: 1.52.2;
Add support "interrupts-extended".

* fdtbus_get_phandle_with_data().
Add utility subroutine to get phandle with data.

Merge isaki-audio2 branch, the overhaul of audio subsystem.
- Interrupt-oriented system rather than thread-oriented.
- Improve stability, quality and performance.
- Split playback and record cleanly. Improve halfduplex support.
- Many bugs are fixed including deadlocks, resource leaks, abuses, etc.
- Simplify audio filter mechanism. The encoding/channels/frequency
conversions are completely handled in the upper layer. So the hard-
ware driver only converts its hardware encoding (if necessary).
- audio_hw_if changes:
- Obsoletes query_encoding and add query_format instead.
- Obsoletes set_params and add set_format instead.
- Remove drain, setfd, mappage.
- The call sequences are changed.
- ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted.
- ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced.
- cleanup config attributes: au*conv and mulaw.
- All hardware drivers should follow it (I've done as much as possible).

Some file paths are changed:
- dev/audio.c -> dev/audio/audio.c (rewritten)
- dev/audiovar.h -> dev/audio/audiovar.h
- dev/audio_dai.h -> dev/audio/audio_dai.h
- dev/audio_if.h -> dev/audio/audio_if.h
- dev/audiobell.c -> dev/audio/audiobell.c
- dev/audiobellvar.h -> dev/audio/audiobellvar.h
- dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c

branches: 1.50.2;
Add fdtbus_intr_establish_byname() helper.

Consolidate FDT [find "names" index by string] logic.

Add support for Allwinner A64's display pipeline.

Define constants for representing the standard interrupt types
({pos,neg,double}-edge, {high,low}-level) from the FDT "interrupts"
bindings. Use these defined constants rather than magic numbers.

Implement subroutines for parsing out some of the generic properties
specified in the pinctrl bindings, and adapt Meson, Rockchip, and
Allwinner pinctrl back-ends to use them.

Ok jmcneill@

Add a helper to search the tree for a node with the specified property

Add a function to report if a regulator can support the requested voltage range

Do not sort cpu nodes when enumerating so they attach in the order listed
in the devicetree.

fdtvar.h: Use _KERNEL_OPT wrapper where necessary.

Add fdt_add_bus_match, which is like fdt_add_bus but allows for the caller to filter child nodes on their own

Pass clock provider's phandle to fdtbus_clock_controller_func.decode()
and update callers.

This allows to accomodate clock managers whose clocks are identified
directly by a clock instead of a pair (clock provider, index).

ok jmcneill@ on port-arm

Add fdtbus_intr_establish_raw and fdtbus_intr_str_raw, for establishing
interrupts directly using an interrupt controller's phandle and specifier.

Add fdtbus_attach_i2cbus, used to attach an i2cbus to an fdt device node.

Add helper to lookup syscon by phandle

Allow nodes to attach child devices using the same sorting rules as used
by the simplebus driver.

Add support for system controller nodes.

branches: 1.34.2;
Process assigned clock parents and rates on clock provider nodes.

Add fdtbus_clock_byname, which can be used by clock backends to
lookup clocks in other domains by "clock-output-names" property.
Not intended for ordinary driver use.

Build fix for FDT kernels without audio

Add SoC sound driver based on "simple-audio-card" DT binding spec.

Add support for PWM backlights.

Allow to remove entries from the fdt, either by handle or
by compatible string. Proposed on tech-kern and port-arm, ok jmcneill@

branches: 1.28.2;
FDTise RapberryPI support. Thanks for jmcneill for a lot of help with this.

The kernel image that the RPI firmware boots is now netbsd.img in the
kernel build directory.

XXX fdtbus_get_reg needs reworking

branches: 1.27.2;
Add support for simple MMC power sequence provider bindings.

Add fdtbus_get_reg_byname

Add fdtbus_gpio_acquire_index for accessing multi-xref gpios properties.

Add fdtbus_get_string_index helper.

Fix the pinctrl api to match the spec. A pinctrl config can have more
than one xref, and an xref may have specifier data associated with it.

Add fdtbus_get_string helper

Add FDT PHY interface.

branches: 1.20.2;
Allow platform code to specify the UART frequency for consinit

Move console initialization out of platform code into the console drivers
themselves.

Add facility for fdt devices to register reset and poweroff handlers.
Use this to remove as3722-specific code in tegra_machdep.c

Handle reg values beyond 4GB for platforms with 32-bit bus_addr_t.

fdtbus_get_reg will do a range check and return ERANGE if addr+size
exceeds 4GB.

Add fdtbus_get_reg64 to allow for fetching raw values w/o range check.

Rename fdt_print to fdtbus_print and make it public.

Move logic for checking "status" property out of fdtbus into a public
fdtbus_status_okay function.

Add FDT DMA controller API.

branches: 1.13.2;
Allow for specifying (optional) pass number locator for devices at fdt.

Add regulator APIs for setting and getting voltage.

Add fdtbus_todr_attach.

The kernel will only ever use the first RTC driver to call todr_attach.
When drivers use fdtbus_todr_attach, if an "rtc0" alias exists in the
devicetree, ensure that only that device node calls todr_attach.

Add some helpers to find the stdout device.

Attach devices in order of "phandle" property inserted by DTC.

Print the full DT node path in "not configured" autoconf messages.

Before: clock at fdt2 not configured
After: /clocks/clock@0 at fdt2 not configured

branches: 1.7.2; 1.7.4;
FDT: Interrupts -- add support for interrupt maps

The mct on exynos uses an interrupt map so we add support now. Devices
represent their interrupts either through a combination of interrupt-parent
and interrupts properties, where the 'interrupts' property is an array of
one or more interrupt specifiers; or through a combination of an
interrupt-parent that points to an interrupt-map, where the interrupt-map
contains 2 or more entries consisting of an index, a pointer to an
interrupt-controller, and a specifier for that controller.

This code adds the ability to walk the interrupt-map and return a specifier.
Unfortunately, the addition requires changing the interface to the
interrupt-controllers' _establish and _intstr functions, so this check in
contains a rototill of the three existing fdt interrupt controllers to use
the new interface.

FDT pinctl - review from Jared

These changes reflect a redesign based on a preliminary review by Jared.
Instead of the acquire/release/set/get approach of the original, this uses
a much simpler, and cleaner register/set approach.

FDT pinctrl

Add a pinctrl bus to FDT. This works against exynos, but someone(tm) needs
to think about whether it is general enough or too specific to exynos.

branches: 1.4.2;
Add fdtbus_gpio_{read,write}_raw, which tells the controller not to take
polarity into account. Tegra GPIO pin data includes pin polarity, but so
does a regulator-fixed node, so the end result was that the enable value
was being swapped twice. Change fregulator to use the raw APIs, and adapt
Tegra and Exynos GPIO drivers to support this flag.

Add support for fdt clock and reset controllers.

Only OF_* functions should be in fdt_openfirm.c, move the rest to fdt_subr.c.

Add a framework for enumerating devices using a Flattened Device Tree (FDT).

Re-factor the console-related code into fdt_console.[ch]

Refactor the "platform" defitions into fdt_platform.h

Step towards modularizing the Flattened Device Tree code.

Define attributes for each of the specific device bindings: clock,
dai, dma, gpio, i2c, iommu, mbox, mmc_pwrseq, phy, power, power domain,
pwm, regulator, reset controller, spi, system controller, pin
controller. Include these support files only if either a provider
or consumer with one of these attributes is present in the kernel
config.

Add the necessary attributes to the device / attach declarations for
each provider and consumer.

There are some bindings that are consumed by generic code (iommu, pinctrl,
power, power domain). Provide weak stubs for these routines to handle
situations where there is no provider.

No actual code changed; NFCI.

branches: 1.73.4;
port-arm/58194: Resurrect vmt(4) from bitrot

On this architecture vmt(4) used to search for a node "/hypervisor" in the
FDT and probed the VMware hypervisor call only when the node was
found. However, things appear to have changed and VMware no longer provides
the FDT node.

Since vmt(4) doesn't actually need to read anything from FDT, and the
hypervisor call logically resides in virtual CPUs themselves, it would be
better to attach it directly to cpu, just like how it's probed on x86.

risc-v: attach the Cadence XHCI usb controller on the JH7100 SoC

Re-factor the Goldfish RTC driver into attach-front-end and generic
back-end; Goldfish virtual devices can be found on virtual platforms
that don't use FDT.

Google goldfish config goop

Move fdt_update_stdout_path from evbarm code to MI code

Separate DesignWare watchdog driver and FDT glue.

Rename ARM_PLATFORM to FDT_PLATFORM and make it available outside arm.

Add driver for Cadence I2C controller.

drm: Use CPPFLAGS.drmkms in all local drm drivers too.

This way we don't pollute the NetBSD kernel namespace with all the
Linux compat shim definitions needed to build drm, except for the
local drm drivers that need the API.

fdt: add support for USERCONF option

Add mailbox and power-domain support.

power-domain support is very rudimentary.

Add facility for registering iommus.

branches: 1.61.10;
simplebus, cpus, and syscon do not need their own interface attribute; they
use the "fdt" interface attribute to attach children.

branches: 1.60.4;
add fdt pwm regulator

Re-factor the "iicmux" driver to support other types of muxes that are
themselves i2c devices.

Add support for i2c-mux-gpio and i2c-mux-pinctrl I2C mux virtual
controllers.

Move evbarm/fdt/fdt_memory.[ch] to sys/dev/fdt and simplify the api and
some operations. This allows other architectures to use it.

branches: 1.56.2;
Add a "show fdt" ddb command

move vmt(4) from MD to MI, and add support vmt on aarch64. tested on ESXi-Arm Fling

- move from sys/arch/x86/x86/{vmt.c,vmtreg.h,vmtvar.h} to sys/dev/vmt/{vmt_subr.c,vmtreg.h,vmtvar.h},
and split the attach part of the cpufeaturebus and fdt
- add aarch64 vmware backdoor op
- add include guard to vmt{reg,var}.h
- Yet there is still some little-endian dependency. it needs to be fixed in order to work properly on aarch64eb